Field of the Invention
The present invention relates to an image processing apparatus and image processing method adapted to extract a defect (unique portion) of an inspection target.
Description of the Related Art
Japanese Patent Laid-Open No. 2013-185862 or ““KIZUKI” Algorithm inspired by Peripheral Vision and Involuntary Eye Movement”, Journal of the Japan Society for Precision Engineering, Vol. 79, No. 11, 2013, p. 1045-1049 discloses an algorism for detecting a defect of an inspection target in accordance with a human visual mechanism. Specifically, after imaging an inspection target, the resulting image is divided into division areas having a predetermined size, and averaging and quantization are performed for each of the division areas. In addition, under multiple sets of conditions determined by changing the size and/or phase of a division area, such a process is performed, and from a result of adding the resulting quantized values, the presence or absence of a defect is determined, and if present, the size of the defect is also determined. By employing such a method, a defect of an inspection target can be efficiently extracted and/or made apparent without any human observation.
When employing the algorithm disclosed in Japanese Patent Laid-Open No. 2013-185862 or the above-described Journal of the Japan Society for Precision Engineering, in order to effectively detect a defect portion (unique portion), it is preferable to adjust various parameters such as a division size and a phase used for an image process. For example, when in order to detect a stripe or unevenness specific to an inkjet printing apparatus, using the above-describe algorithm to inspect an image printed by the printing apparatus, a division size range and a phase range for suitably detecting the stripe or unevenness change depending on the features of the stripe or unevenness. In addition, when such parameters are not properly adjusted, the defect portion may be unextractable, or a load on and time for the extracting process may be uselessly increased.
However, in Japanese Patent Laid-Open No. 2013-185862 or the above-described Journal of the Japan Society for Precision Engineering, the correspondence relationship between the features of a defect to be extracted and parameters suitable for the defect extraction has not been referred to at all.